Chest roentgenography and sputum cytology have been used for the early detection of lung cancer since the beginning of the 1950s. Surprisingly, the yield of screening is sufficiently high. In view of the estimated prevalence of lung cancer of 0.5-2% in high risk persons, the detection rate was 0.32-0.95% in three large randomized clinical trials (Cooperative Pilot Study of the American Cancer Society and Veterans Administration 1969, Canadian study of Grzybowski and Coy 1970, National Cancer Institute Cooperative Early Lung Cancer Detection Program 1984). Screening by chest roentgenogram was superior to sputum cytology. More than 70% of all lung cancers were detected by roentgenography alone. Adding sputum cytology, 10-20% more cancers were found. Lung cancers which were additionally detected by cytology were slowly-growing, central squamous cell carcinomas with good prognosis. The sensitivity of chest X-rays varied between 49 and 77% and the sensitivity of sputum cytology between 33-52%. Specificity was higher, the range being 95-99% for chest X-rays and 96-99% for sputum cytology. The costs per lung cancer detected and treated are very high. To these are added the costs which arise from the evaluation of false positive results. The prognosis of lung cancer detected by screening is good: more patients survive, and they live longer than unscreened patients. The overall 5-year survival of patients screened by chest X-rays and sputum cytology of the NCI-study was 35-55% compared with only 15% of patients not screened. However, whether long survival means reduced mortality is uncertain, since long survival only reflects early detection.